A high-reliability anti-interference engine rotating speed measurement data processing method
By generating interval time series based on square wave time series and combining multi-order difference and smoothing filtering, the noise and interference problems in engine speed measurement are solved, and the accuracy and reliability of speed measurement are improved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SICHUAN AEROSPACE ZHONGTIAN POWER EQUIP CO LTD
- Filing Date
- 2023-03-24
- Publication Date
- 2026-05-29
AI Technical Summary
Existing technologies for engine speed measurement suffer from a large amount of high-frequency noise, random interference, and intermittent or continuous sampling anomalies, resulting in problems such as the loss of a lot of valid data, large prediction deviations, and speed jumps. There is a lack of effective methods to reasonably divide and classify the time intervals collected by speed sensors.
Square wave time series are generated based on square wave time series. Rotation speed is calculated by multivariate tuple of the interval time series. Outliers are checked by multi-order difference and multi-point prediction methods. Finally, smoothing filtering is performed to generate optimized rotation speed.
It enables reasonable division and classification of rotational speed under high-frequency noise and random interference, reduces the loss of effective data, improves measurement accuracy, suppresses speed jumps, and calculates more reliable real-time rotational speed.
Smart Images
Figure QLYQS_5 
Figure QLYQS_7 
Figure QLYQS_11
Abstract
Description
Technical Field
[0001] This invention relates to the field of engine speed measurement technology, and specifically to a highly reliable and interference-resistant engine speed measurement data processing method. Background Technology
[0002] The entire operation of an aero-engine is a highly complex aero-thermodynamic process. Due to the significant differences in aero-thermodynamic processes under various environmental conditions and states, the mathematical models of aero-engine systems typically exhibit strong nonlinear characteristics, leading to nonlinear variations in engine speed. Engine speed is a crucial parameter for engine performance and a key target parameter for engine control. Engine speed measurement data is affected by factors such as limited sensor installation space, a large measurement speed range, high-frequency mechanical vibration, electromagnetic interference, and sensor drift, resulting in significant high-frequency noise, random interference, and intermittent or continuous sampling anomalies.
[0003] Traditional methods primarily employ mean filtering, differential filtering, and prediction to smooth and predict the calculated measured rotational speed. However, these methods suffer from issues such as significant data loss, large prediction errors, and rotational speed jumps. There is a lack of effective methods to directly and rationally classify the time intervals collected by the rotational speed sensor, thereby reducing data loss, improving measurement accuracy, suppressing speed jumps, and calculating a more reasonable and reliable real-time rotational speed. Summary of the Invention
[0004] The technical problem to be solved by this invention is: how to design a highly reliable and anti-interference engine speed measurement data processing method that can reasonably divide and classify the time intervals collected by the speed sensor, reduce the loss of effective data, improve the speed measurement accuracy, suppress speed jumps, and calculate a more reasonable and reliable real-time speed under conditions such as a large amount of high-frequency noise, random interference, and intermittent or continuous time sampling anomalies.
[0005] To achieve the above objectives, the present invention adopts the following technical solution:
[0006] A highly reliable and interference-resistant engine speed measurement data processing method includes the following steps:
[0007] A square wave interval time series is generated based on the square wave time series, and an interval time series tuple is generated based on the square wave interval time series. The rotational speed is then calculated using the interval time series tuple.
[0008] Check and replace outlier values in the calculated rotational speed to obtain an optimized rotational speed;
[0009] The optimized speed is smoothed and filtered to obtain the output speed.
[0010] In some embodiments, the step of generating a square wave interval time series based on a square wave time series, generating an interval time series tuple based on the square wave interval time series, and obtaining the calculated rotational speed through the interval time series tuple includes:
[0011] Square wave time series are generated by acquiring square wave time series using an eddy current speed measurement system. The square wave time series is then optimized to generate a tuple of interval time series. The calculated speed is obtained from the tuple of interval time series.
[0012] In some embodiments, a square wave interval time series is generated using a square wave time series acquired by an eddy current speed measurement system, the square wave interval time series is optimized to generate an interval time series tuple, and the calculated speed is obtained from the interval time series tuple. Specifically, this includes:
[0013] The speed measurement system processes the changing voltage signal into a square wave signal, detects the falling edge of the square wave and records the time T at which the falling edge occurs, thus forming a square wave time series. ; Calculate the time difference between two adjacent moments in the square wave time series to generate a square wave interval time series. ,in .
[0014] In some embodiments, obtaining the calculated rotational speed specifically includes the following steps:
[0015] S1: Optimize the square wave interval time series Generate interval time series tuples Where m represents the tuple index of the time series interval, and i and j represent the number of elements in the tuple, respectively;
[0016] S2: Calculate the rotational speed by using a tuple of time series intervals.
[0017] In some embodiments, S1: Optimize the square wave interval time series Generate interval time series tuples The specific steps include:
[0018] S10: Calculate the square wave interval time series mean With variance :
[0019] , in Square wave interval time series The i-th element in the sequence, where n is the total number of points in the square wave interval time series;
[0020] choose As the initial point , = The interval time series tuple number m=1;
[0021] S11: From the square wave interval time series Starting from the initial point, iterate to the last point of the square wave interval time series. ;
[0022] If the initial point is satisfied The last point of the square wave interval time series The sequence number then generates an interval time series tuple. ;
[0023] If satisfied ,in For the mean, If the variance is given, then from the initial point... To the end point Generate interval time series tuples ;
[0024] If satisfied And it iterates to the last point of the square wave interval time series. That is, j=n, where For the mean, If the variance is given, then from the initial point... To the end point Generate interval time series tuples ;
[0025] S12: If the index of the last point in the interval time series tuple generated by S11 is less than the last point in the square wave interval time series... If the index is given, then let the initial point be... =Increment 1 by the index of the last point in the interval time series tuple, where the interval time series tuple index is... Repeat S11 until the condition is met.
[0026] In some embodiments, S2: obtaining the calculated rotational speed through an interval time series tuple includes the following steps:
[0027] S20: Calculate the total interval time for each interval time series tuple to generate the total interval time series. ,in , where i is the index of the i-th interval time series tuple, and j is the number of points in the i-th interval time series tuple. Let j be the j-th point in the i-th interval time series tuple;
[0028] S21: The unit for calculating the rotational speed is "revolutions per minute" (rpm), which is related to the number of engine impeller blades (p) and the interval time; based on the total interval time sequence... The calculated rotational speed is obtained by calculating the number of impeller blades p. ,in Redirects / minutes.
[0029] In some embodiments, the step of checking and replacing outlier values in the calculated rotational speed to obtain an optimized rotational speed includes: checking the calculated rotational speed... A multi-order difference and multi-point prediction method is used to check and replace outlier values in the calculated rotational speed to obtain an optimized rotational speed.
[0030] In some embodiments, the calculation of rotational speed The following steps were taken to check and replace outliers in the calculated rotational speed using a multi-order difference and multi-point prediction method to obtain the optimized rotational speed:
[0031] S30: Use multi-order difference pairs to calculate and remove abnormal speed values;
[0032] Calculate each rotational speed The corresponding 5th order difference value ,in This belongs to the calculation of rotational speed ;
[0033] If the 5th order difference value ,in for variance, then Let be a set of reasonable points; otherwise, let Continue with differential verification;
[0034] S31: Use a multi-point forecasting method to check and replace outliers in the calculation;
[0035] Select a set of reasonable points that have been filtered in S30 ,calculate linear forecast value ,in
[0036] ;
[0037] if , where σ is The variance, then To achieve a reasonable value, optimize the rotational speed. ;if , where σ is The variance, then The speed is unreasonable; optimize it. ;
[0038] S32: Order Repeat steps S30 and S31 to find a set of reasonable points. Use a multi-point forecasting method to check and replace outliers in the calculated rotation speed to generate an optimized rotation speed. .
[0039] In some embodiments, the step of performing smoothing filtering on the optimized rotational speed to obtain the output rotational speed specifically includes:
[0040] The optimized speed is smoothed and filtered to obtain the output speed. The following steps are used:
[0041] S41: A second-order filtering method is used to optimize the speed filtering. ,in Let be the filter coefficients, satisfying , , ;
[0042] S42: Traversal optimization speed For each optimized speed, repeat the calculation in S41 to calculate the output speed. .
[0043] The highly reliable and interference-resistant engine speed measurement data processing method provided in this application has the following beneficial effects, including but not limited to:
[0044] This invention utilizes a square wave time series acquired by an eddy current speed measurement system to generate a square wave interval time series. The square wave interval time series is then optimized to generate a multivariate set of interval time series, from which the calculated speed is obtained. A multi-order difference and multi-point prediction method is employed to check and replace outliers in the calculated speed, resulting in an optimized speed. This optimized speed is then smoothed and filtered to obtain the output speed. This method effectively removes interference signals, high-frequency signals, and abnormal data, improving the reliability and data quality of the output speed. It exhibits strong anti-interference, anti-drift, and good real-time performance. Detailed Implementation
[0045] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.
[0046] Conversely, this application covers any substitutions, modifications, equivalent methods, and schemes made within the spirit and scope of this application as defined in the claims. Furthermore, to provide the public with a better understanding of this application, certain specific details are described in detail below. However, this application can be fully understood by those skilled in the art even without these detailed descriptions.
[0047] The following will provide a detailed description of a highly reliable and interference-resistant engine speed measurement data processing method according to embodiments of this application. It is worth noting that the following embodiments are merely illustrative of this application and do not constitute a limitation thereof.
[0048] A highly reliable and interference-resistant engine speed measurement data processing method includes the following steps:
[0049] Step 1: Generate a square wave interval time series using the square wave time series acquired by the eddy current speed measurement system, optimize the square wave interval time series to generate an interval time series tuple, and obtain the calculated speed through the interval time series tuple.
[0050] Step 2: Use multi-order difference and multi-point prediction methods to check and replace outlier values in the calculated rotational speed to obtain the optimized rotational speed.
[0051] Step 3: Perform smoothing filtering on the optimized speed to obtain the output speed.
[0052] Step 1 specifically includes:
[0053] When the centrifugal impeller of an engine rotates, an eddy current is induced in the impeller and forms an eddy current in the speed measurement system. This eddy current causes a change in the impedance of the speed measurement system, resulting in a change in voltage. The speed measurement system processes this changing voltage signal into a square wave signal, detects the falling edge of the square wave, and records the time T at which this edge occurs, forming a square wave time series. The time difference between two adjacent moments in a square wave time series is calculated to generate a square wave interval time series. ,in Based on the above principles, obtaining the calculated rotational speed mainly includes the following steps:
[0054] (1) Optimize the square wave interval time series Generate interval time series tuples , where m represents the tuple index of the time series interval, and i and j represent the number of elements in the tuple, respectively.
[0055] (2) The rotational speed is obtained by using a tuple of time intervals.
[0056] Optimize square wave interval time series Generate interval time series tuples The specific steps include:
[0057] Step 1: Calculate the square wave interval time series mean With variance :
[0058] , in Square wave interval time series The i-th element in the sequence, where n is the total number of points in the square wave interval time series.
[0059] choose As the initial point , = The interval time series tuple number m=1.
[0060] Step Two: From the square wave interval time series Starting from the initial point, iterate to the last point of the square wave interval time series. .
[0061] If the initial point is satisfied The last point of the square wave interval time series The sequence number then generates an interval time series tuple. .
[0062] If satisfied ,in For the mean, If the variance is given, then from the initial point... To the end point Generate interval time series tuples .
[0063] If satisfied And it iterates to the last point of the square wave interval time series. That is, j=n, where For the mean, If the variance is given, then from the initial point... To the end point Generate interval time series tuples .
[0064] Step 3: If the index of the last point in the interval time series tuple generated in step 2 is less than the last point in the square wave interval time series... If the index is given, then let the initial point be... =Increment 1 by the index of the last point in the interval time series tuple, where the interval time series tuple index is... Repeat step two.
[0065] The steps to obtain the calculated rotational speed using interval time series tuples are as follows:
[0066] Step 1: Calculate the total interval time for each interval time series tuple to generate the total interval time series. ,in , where i is the index of the i-th interval time series tuple, and j is the number of points in the i-th interval time series tuple. Let j be the j-th point in the i-th interval time series tuple.
[0067] Step 2: Calculate the rotational speed in revolutions per minute (rpm), which is related to the number of engine impeller blades (p) and the interval time. Based on the total interval time sequence... The calculated rotational speed is obtained by calculating the number of impeller blades p. ,in Redirects / minutes.
[0068] Step 2 specifically includes:
[0069] For calculating the rotational speed The multi-order difference and multi-point prediction method is used to check and replace outliers in the calculated speed to obtain the optimized speed. The steps are as follows:
[0070] Step 1: Use multi-order difference pairs to calculate and check for and eliminate abnormal speed values.
[0071] Calculate each rotational speed The corresponding 5th order difference value ,in This belongs to the calculation of rotational speed If the 5th order difference value ,in for The variance, then Let be a set of reasonable points; otherwise, let Continue with differential verification.
[0072] Step 2: Use a multi-point forecasting method to check and replace outliers in the calculation. Select a set of reasonable points already screened in Step 1. ,calculate linear forecast value ,in .if ,in for The variance, then To achieve a reasonable value, optimize the rotational speed. .if ,in for variance, then The speed is unreasonable; optimize it. .
[0073] Step 3: Order Repeat steps one and two to find a set of reasonable points. Use a multi-point forecasting method to check and replace outliers in the calculated rotation speed, and generate an optimized rotation speed. .
[0074] Step 3 specifically includes:
[0075] The optimized speed is smoothed and filtered to obtain the output speed. The following steps are used:
[0076] Step 1: Optimize the speed filtering using a second-order filtering method. ,in Let be the filter coefficients, satisfying , , .
[0077] Step 2: Traverse and optimize rotation speed For each optimized speed, repeat the calculation in step one to calculate the output speed. .
[0078] This invention relates to a highly reliable and interference-resistant engine speed measurement data processing method. The method utilizes a square wave time series acquired by an eddy current speed measurement system to generate a square wave interval time series. This interval time series is then optimized to generate a multivariate set of interval time series, from which the calculated speed is obtained. A multi-order difference and multi-point prediction method is employed to check and replace outliers in the calculated speed, resulting in an optimized speed. This optimized speed is then smoothed and filtered to obtain the output speed. This method effectively removes interference signals, high-frequency signals, and outlier data, improving the reliability and quality of the output speed. It exhibits strong anti-interference, anti-drift, and good real-time performance.
[0079] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A highly reliable and interference-resistant engine speed measurement data processing method, characterized in that, Includes the following steps: A square wave interval time series is generated based on the square wave time series, and an interval time series tuple is generated based on the square wave interval time series. The rotational speed is then calculated using the interval time series tuple. include: Square wave time series are generated from square wave time series acquired by an eddy current speed measurement system. The square wave time series is then optimized to generate a tuple of interval time series. The speed is calculated using this tuple of interval time series, specifically including: The speed measurement system processes the changing voltage signal into a square wave signal, detects the falling edge of the square wave and records the time T at which the falling edge occurs, thus forming a square wave time series. ; Calculate the time difference between two adjacent moments in the square wave time series to generate a square wave interval time series. ,in ; To obtain the calculated rotational speed, the following steps are involved: S1: Optimize the square wave interval time series Generate interval time series tuples Where m represents the tuple index of the time series interval, and i and j represent the number of elements in the tuple, respectively; S2: Calculate the rotational speed by using a tuple of time-series intervals; S1: Optimize the square wave interval time sequence Generate interval time series tuples The specific steps include: S10: Calculate the square wave interval time series mean With variance : , Where t i Square wave interval time series The i-th element in the sequence, where n is the total number of points in the square wave interval time series; choose As the initial point , = The interval time series tuple number m=1; S11: From the square wave interval time series Starting from the initial point, iterate to the last point of the square wave interval time series. ; If the initial point is satisfied The last point of the square wave interval time series The sequence number then generates an interval time series tuple. ; If satisfied ,in For the mean, If the variance is given, then from the initial point... To the end point Generate interval time series tuples ; If satisfied And it iterates to the last point of the square wave interval time series. That is, j=n, where For the mean, If the variance is given, then from the initial point... To the end point Generate interval time series tuples ; S12: If the index of the last point in the interval time series tuple generated by S11 is less than the last point in the square wave interval time series... If the index is given, then let the initial point be... =Increment 1 by the index of the last point in the interval time series tuple, where the interval time series tuple index is... Repeat S11 until the condition is met. S2: Obtaining the calculated rotational speed through interval time series tuples includes the following steps: S20: Calculate the total interval time for each interval time series tuple to generate the total interval time series. ,in where i is the index of the i-th interval time series tuple, and j is the number of points in the i-th interval time series tuple. Let j be the j-th point in the i-th interval time series tuple; S21: The unit for calculating the rotational speed is "revolutions per minute" (rpm), which is related to the number of engine impeller blades (p) and the interval time; based on the total interval time sequence... The calculated rotational speed is obtained by calculating the number of impeller blades p. ,in Turns / minute; Check and replace outlier values in the calculated rotational speed to obtain an optimized rotational speed; The optimized speed is smoothed and filtered to obtain the output speed.
2. The highly reliable and interference-resistant engine speed measurement data processing method according to claim 1, characterized in that, The process of checking and replacing outliers in the calculated rotational speed to obtain an optimized rotational speed includes: checking the calculated rotational speed... A multi-order difference and multi-point prediction method is used to check and replace outlier values in the calculated rotational speed to obtain an optimized rotational speed.
3. The highly reliable and interference-resistant engine speed measurement data processing method according to claim 2, characterized in that, The calculation of rotational speed The following steps were taken to check and replace outliers in the calculated rotational speed using a multi-order difference and multi-point prediction method to obtain the optimized rotational speed: S30: Use multi-order difference pairs to calculate and remove outlier speed values; Calculate each rotational speed The corresponding 5th order difference value ,in This belongs to the calculation of rotational speed ; If the 5th order difference value ,in for The variance, then Let be a set of reasonable points; otherwise, let Continue with differential verification; S31: Use a multi-point forecasting method to check and replace outliers in the calculated rotational speed; Select a set of reasonable points that have been filtered in S30 ,calculate linear forecast value ,in ； if , where σ is The variance, then To achieve a reasonable value, optimize the rotational speed. ;if , where σ is The variance, then The speed is unreasonable; optimize it. ; S32: Order Repeat steps S30 and S31 to find a set of reasonable points. Use a multi-point forecasting method to check and replace outliers in the calculated rotation speed to generate an optimized rotation speed. .
4. The highly reliable and interference-resistant engine speed measurement data processing method according to claim 1, characterized in that, The process of smoothing and filtering the optimized rotational speed to obtain the output rotational speed specifically includes: The optimized speed is smoothed and filtered to obtain the output speed. The following steps are used: S41: A second-order filtering method is used to optimize the speed filtering. ,in Let be the filter coefficients, satisfying , , ; S42: Traversal optimization speed For each optimized speed, repeat the calculation in S41 to calculate the output speed. .